9999.99 = N.I.U.

1982-1988:

• 999 (Weekly earnings of $999 or more)

1989-1997:

• 1923 (Weekly earnings of $1923 or more)

1998-March 2023:

• ASEC samples only: 2885 (Weekly earnings of $2885 or more)

• non-ASEC samples: 2884.61

April 2023-onward:

• MISH 4: The maximum value of EARNWEEK2, excluding NIU values

• MISH 8: 2884.61

Description

EARNWEEK2 reports how much the respondent usually earned per week at their current job, before deductions, rounded to preserve privacy. In April 2023, the Census Bureau began rounding weekly earnings as a privacy protection measure. EARNWEEK2 imposes these rounding criteria on pre-April 2023 data (including the 2023 ASEC) to provide users with comparable weekly earnings over time. See the Census Bureau's user note for more information on the new privacy protection measures. EARNWEEK reports un-rounded weekly earnings for pre-April 2023 data (including the 2023 ASEC) only.

Note that rounding of earnings is part of a set of changes made to preserve respondent privacy that began to be phased in with the MISH group in January of 2023. April of 2023 is the first month in which this MISH group is asked earnings questions. Because of Census Bureau's phased approach to introducing these privacy protection measures, only MISH 4 groups in April 2023 - March of 2024 Basic Monthly samples are rounded and topcoded by the Census Bureau. In these months, IPUMS applies the new rounding rules to the MISH 8 group so that all records have comparable values within months. Users wishing to access a variable in which MISH 4 is rounded and MISH 8 remains unrounded should use UH_ERNWK_B5.

Respondents reporting earnings between $0.01 and $7 have weekly earnings set at $4. Those who report $7.01-$1,199.00 have their weekly earnings values rounded to the nearest $2. Weekly earnings between $1,199.01 and $1,999.00 are rounded to the nearest $10. Finally, weekly earnings at or above $2,000.00 are rounded to the nearest $20. For further information about the changes to earnings data beginning in April 2023 and rounding examples, see the Census Bureau's user note.

Interviewers asked directly about total weekly earnings and collected information about the usual number of hours worked per week and the hourly rate of pay at the current job. The figure given in EARNWEEK2 is the higher of the values derived from these two sources: 1) the respondent's answer to the question, "How much do you usually earn per week at this job before deductions?" or 2) for workers paid by the hour (and coded as "2" in PAIDHOUR), the reported number of hours the respondent usually worked at the job, multiplied by the hourly wage rate given in HOURWAGE, and rounded as described above.

Amounts are expressed as they were reported to the interviewer, users must adjust for inflation using Consumer Price Index adjustment factors. Researchers should use the EARNWT weight with this variable.

EARNWEEK2 is one of the Outgoing Rotation/Earner Study questions.

Comparability

Beginning in April 2023, the Census Bureau instituted a "dynamic" topcode for weekly earnings in which the top three percent of cases are topcoded and assigned a value that is a weighted average of the weekly earnings values of all cases that fall within that top three percent. As a result, topcode values provided by the Census Bureau change monthly for the MISH 4 group from April 2023 - March 2024 and for all respondents in the outgoing rotation groups from April 2024 onward. Prior to April 2023, the topcode for weekly earnings represented four to five percent of the values while the dynamic topcoded samples have only 3% of values are topcoded. IPUMS does not round original topcoded values when applying new rounding rules to originally un-rounded months or rotation groups. The Census Bureau recommends using the lowest topcode for the months of interest. For further information about topcoding, see the Census Bureau’s user note.

EARNWEEK2 as provided by IPUMS has 2 decimal places. Users should note that EARNWEEK2 originally had no implied decimals in any ASEC samples, nor in non-ASEC samples from 1982-1993. There are 2 implied decimals in non-ASEC samples from 1994 onwards. Command files provided by IPUMS multiply EARNWEEK2 by 100 in samples without implied decimals to harmonize the difference, so no further adjustment is necessary.

Information about weekly earnings was collected beginning in 1982, as part of a series of questions about usual weekly and hourly earnings. The Census Bureau reports that the results in the CPS public use files for this data series included errors for ASEC files prior to 1989. As a result, these variables are only available in ASEC data through IPUMS CPS from 1989 forward. Monthly data is available from 1982 forward.

In the original data CPS data from 1982-1988, weekly earnings are split into separate variables for those who are paid hourly and those who are not. In order to facilitate comparability with later samples, IPUMS CPS has combined information from both of these variables into EARNWEEK2 during this period.

According to Technical Paper 66, issued jointly by the Census Bureau and the Bureau of Labor Statistics, individuals eligible for the earner study are civilians age 15 and older in rotation groups 4 or 8 who are not self-employed.

This is true for data from 1989 forward, but in the data itself, the universe for 1982 to 1988 is civilians age 14 and older in rotation groups 4 or 8 who are not self-employed.

In any given month, approximately 1/4 of the CPS sample is in the earner study and each household should appear in the earner study exactly twice. Based on documentation from Unicon and NBER, and after an inspection of the original CPS data, we recommend that users impose the CPS eligibility restrictions in any analyses of earner study variables.

Universe

Flags

Unharmonized Variables

• UH_ERNWK_A2
• UH_ERNWK_A3
• UH_ERNWK_B1
• UH_ERNWK_B2
• UH_ERNWK_B3
• UH_ERNWK_B4
• UH_ERNWK_B5
• UH_ZERNWK_A2